Experimental Neurology 253 (2014) 165–173

Contents lists available at ScienceDirect

Experimental Neurology

j ourna l homepage: www.e lsev ie r .com/ locate /yexnr

Treatment with anti-TNF alpha protects against the neuropathy inducedby the proteasome inhibitor bortezomib in a mouse model

Albert Alé a, Jordi Bruna a, Marta Morell a, Helgi van de Velde b, Johan Monbaliu b,Xavier Navarro a, Esther Udina a,⁎a Institute of Neurosciences and Department of Cell Biology, Physiology and Immunology, Universitat Autònoma de Barcelona, and Centro de Investigación Biomédica en Red sobre EnfermedadesNeurodegenerativas (CIBERNED), Spainb Janssen Research and Development, Beerse, Belgium

⁎ Corresponding author at: Faculty of Medicine, Unive08193 Bellaterra, Spain. Fax: +34 93 5812986.

E-mail address: esther.udina@uab.cat (E. Udina).

0014-4886/$ – see front matter © 2014 Elsevier Inc. All rihttp://dx.doi.org/10.1016/j.expneurol.2013.12.020

a b s t r a c t

a r t i c l e i n f o

Article history:Received 29 October 2013Revised 10 December 2013Accepted 27 December 2013Available online 7 January 2014

Keywords:BortezomibNeuropathyCytokinesChemotherapy-induced peripheral neuropathyProteasome inhibitorMice

Bortezomib (BTZ), a proteasome inhibitor, is an effective anti-neoplastic drug used in the treatment of multiplemyeloma andmantle cell lymphoma. However, it can induce a reversible peripheral neuropathy that may lead totreatment discontinuation. The mechanism through which BTZ exerts toxic effects in peripheral neurons is notclear. Release of proinflammatory cytokines after nerve damage can induce neurodegeneration, but the effectsof BTZ on cytokine expression in neurons are unknown, although BTZ modulates the expression of cytokines,such as TNF-α and IL-6, in tumor cells. The aim of this study was to evaluate the expression and the roleof these cytokines on the course of BTZ induced neuropathy in mice. IL-6, TNF-α, TGF-β1 and IL-1β wereup-regulated in dorsal root ganglia but TNF-α and IL-6 increased faster and higher. Then, we studied thepotential neuroprotective effect of selective antibodies anti-TNF-α and anti-IL-6 on the evolution of theneuropathy. Treatment with anti-TNF-α but not with anti-IL-6 significantly prevented the decrease ofsensory nerve action potentials amplitude and the loss of myelinated and unmyelinated fibers.We concludethat monoclonal antibodies directed against TNF-α may be a suitable neuroprotective therapy against theneurotoxicity induced by BTZ.

© 2014 Elsevier Inc. All rights reserved.

Introduction

Bortezomib (BTZ), a boronid acid dipeptide, is the first of a new classof chemotherapeutic drugs that inhibit the 20S proteasome complex.BTZ is an effective treatment for multiple myeloma and mantle celllymphoma (Fisher et al., 2006; Richardson et al., 2003; San Miguelet al., 2008). As the main dose limiting toxicity, BTZ induces in themost cases a reversible peripheral neuropathy (Badros et al., 2007;Mateos et al., 2008; Richardson et al., 2003), which is mainly sensory,length dependent, potentially painful and thus, suggestive of involve-ment of Aβ, Aδ and C primary afferent fibers. The neuropathic symp-toms lead to reduce the dosage or even withdraw the treatment,consequently affecting the prognosis and the quality of life of thepatients.

The mechanisms through which BTZ induces the development ofperipheral neuropathy remain poorly understood. Several mechanismshave been proposed to contribute to BTZ-induced peripheral neuropa-thy, including alterations in the axonal cytoskeleton (Poruchynskyet al., 2008; Staff et al., 2013), effects on mRNA translation due to

rsitat Autònoma de Barcelona,

ghts reserved.

nuclear accumulation of ubiquitylated proteins and poly(A)RNAs andDNA damage (Casafont et al., 2010; Palanca et al., 2013), decrease inSchwann cell dedifferentiation (H.K. Lee et al., 2009) and toxic effectson mitochondria in primary afferent sensory neurons (Zheng et al.,2012). In cancer cells, it has been proposed that BTZ induces apoptosisby inhibiting NF-κB, leading to reduced production of interleukin 6(IL-6) and tumor necrosis factorα (TNF-α). These twoproinflammatorycytokines are related to apoptosis and the adhesion capabilities oftumoral cells (Chauhan et al., 1996; Hideshima et al., 2001; Wanget al., 1998). However, the effects of BTZ induced changes in cytokineson neurons are poorly known. Cytokines play a key role in the balanceof cell signals for proliferation, differentiation and apoptosis, and inthe control of responses during nerve degeneration and regeneration(Dubovy, 2011; Myers and Shubayev, 2011). In fact, both IL-6 andTNF-α increase after peripheral nerve damage from trauma or nervetoxic agents as chemotherapy drugs (Myers et al., 2006), and protea-some inhibition can induce proinflammatory response in neuronalcells (Rockwell et al., 2000). In fact, alterations in the inflammatoryprocesses have been related to BIPN in patients (Watanabe et al.,2013), although immunodeficient mice develop similar neuropathythan immunocompetent ones (Carozzi et al., 2013). On the otherhand, the increase of TNF-α is pointed out as a key factor in the develop-ment of neuropathic pain and small fiber pathology after neuronal

166 A. Alé et al. / Experimental Neurology 253 (2014) 165–173

injury (Leung and Cahill, 2010; Myers and Shubayev, 2011). Therefore,in this study we aimed to analyze how BTZ affects the expression ofthese cytokines in the neuronal tissue. They may be downregulatedsimilarly to what is observed in tumor cells or, in contrast, may beupregulated, following a neuroinflammatory response similar to theobserved in other neural disorders.

We have used an animal model previously characterized in ourlaboratory, in which the neuropathy induced by BTZ shows similarfeatures to those described in patients, with decrease of sensory nerveaction potentials (SNAPs), decrease in number of myelinated axons,partial loss of unmyelinated fibers in the skin, and reversibility afterwithdrawal (Bruna et al., 2010). As in humans, we observed selectiveinvolvement of sensory nerve conduction tests without motor conduc-tion abnormalities, suggesting that the toxicity was mostly affectingthe sensory neurons located in the dorsal root ganglia (DRG). Sinceboth cytokines, IL-6 and TNF-α, were found markedly upregulated inDRG after BTZ administration, we also studied the potential therapeuticeffects of the administration of selective antibodies anti-TNF-α andanti-IL-6 on the evolution of the neuropathy.

Material and methods

OF1 female mice aged 2.5 months were used in the study. Theexperimental procedures were approved by the ethics committeeof the Universitat Autònoma de Barcelona, and were carried out inaccordance with the European Community Council Directives.

mRNA analysis of cytokines

A batch of mice treated with BTZ at 1 mg/kg subcutaneously twiceperweekwas used to analyze themRNA expression of cytokines duringthe treatment. Twenty animals were sacrificed by decapitation afterdeep anesthesia at 1, 2, 4 or 6 weeks of treatment. DRG, spinal cordand sciatic nerves were rapidly dissected, maintained in RNA-latersolution (Qiagen, Barcelona, Spain) and processed for mRNA analyses.Total RNA was extracted using RNeasy Mini kit (Qiagen) includinga DNase step (RNase free DNase set, Qiagen). One microgram ofRNA per sample was reverse-transcribed using 10 μmol/l DTT, 200 UM-MuLV reverse transcriptase (New England BioLabs, Barcelona, Spain),10 U RNase Out Ribonuclease Inhibitor (Invitrogen) and l μmol/loligo(dT), 1 μmol/l of random hexamers (BioLabs, Beverly, MA, USA).The reverse transcription cycle conditions were 25 °C for 10 min, 42 °Cfor 1 h and 72 °C for 10 min. We analyzed the mRNA expressionof TNF-α, IL-6, TNF-α receptor 1 (TNFR1), IL-6 receptor α (IL-6Rα),interleukin 1β (IL-1β), and transforming growth factor β1 (TGF-β1).Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) expression wasused to normalize the expression levels of the cytokines (see list ofprimers in Table 1).

Gene-specific mRNA analyses were performed by SYBR-green real-time PCR using the MyiQ5 real-time PCR detection system (BioRadLaboratories, Barcelona, Spain). We previously fixed the optimalconcentration of the cDNA to be used as template for each gene analysisto obtain reliable CT (threshold cycle) values for quantification.Realtime PCR amplification reactions contained the same amount of

Table 1List of primers used for real time PCR analysis in this study. Forward and reverse primersare indicated for each gene.

Primer Forward (5′➔3′) Reverse (5′➔3′)

TNF-α AAAGGGAGAGTGGTCAGG TCTGTGAGGAAGGCTGTGTNFR1 AAAGCCACACCCACAACC ATAAAGCCACACCCACAACCIL-6 ACAAGAAAGACAAAGCCAGAG TCTTGGTCCTTAGCCACTCIL-6Rα TCTGCTTCCTGCTACACCTC AGTAAGACGGCAGGTGTGGIL-1β TTCCCAATCCCTCAACAGTC ATGTTCTGGAGCAGGCAGTGTGF-β1 TGGAGCTGGTGAAACGGAAG ACAGGATCTGGCCACGGATGAPDH AGTTCAACGGCACAGTCAAGGC TACTCAGCACCGGCGTCACC

RT product, 10 μl of 2× QuantiMix EASY SYG KIT (Biotools, Madrid,Spain), 300°nM of forward and reverse primers, and completed withnanopure water to obtain a final volume of 20 μl. The thermal cyclingconditions comprised 3 min polymerase activation at 95 °C, 40°cyclesof 10 s at 95 °C for denaturation and 30 s at 60 °C for annealing andextension, followed by a DNA melting curve for the determination ofamplicon specificity. All experiments were performed in duplicate. CTvalues were obtained and analyzed with the BioRad Software. Foldchange in gene expression was estimated using the CT comparativemethod (2−ΔΔT) normalizing to GAPDH CT values and relative tocontrol samples. The fold-increase expression of samples during BTZtreatment with respect to basal samples was calculated for eachtranscript, and was based on real-time PCR threshold values. Themean value was obtained from three pools of each tissue per conditionin each group.

Protein extraction and Western blot

For protein extraction, 4 control animals and 4mice treatedwith BTZfor one week were anesthetized and decapitated, and DRG and sciaticnerves quickly harvested. Samples were processed for subcellularfractionation to separate the nucleus from the cytoplasm. Protein wasextracted and homogenized in two different solutions. First, membranelysis bufferwas added (10 mMHepes pH 7.4, 8.5% sucrose, 1 mMEGTA,1 mMEDTA, 10 μl/ml Protease Inhibitor cocktail from Sigma and 1 mMsodiumorthovanadate, a phosphatase inhibitor, from Roche). After cen-trifugation at 800 g the supernatant was removed to obtain cytoplasmicproteins. Nucleus lysis buffer was added to the pellet (5 mM HepespH 7.4, 0.1 mM EGTA, 0.1 mM EDTA, 0.2 mM KCl, 0.1% Triton X-100,5 μl/ml Protease Inhibitor cocktail, and 1 mM sodium orthovanadate),and the sample was vortexed 20 min at 4 °C and centrifugated at12,000 g. After clearance, protein concentration was quantified using amodified Lowry assay (Bio-Rad Dc protein assay).

To perform theWestern blot, 20 μg of protein from each samplewasloaded in 7.5% SDS-polyacrylamide gels. The transfer buffer consisted in25 mM trizma-base, 192 mMglycine and 20% (v/v)methanol at pH 8.4.The membranes were incubated with 5% BSA in PBS plus 0.05% Tween-20 for 1°h, and then overnight at 4 °C with primary antibodies, anti-β-actin (Sigma; 1:10,000), anti phosphorylated p65 (Cell Signalling,1:1000), anti anti-p65 (Santa Cruz, 1:200) and anti-IκB-α (Cell Signal-ling, 1:1000). Horseradish peroxidase-coupled secondary antibodyincubation was performed for 90 min at room temperature. The mem-branes were visualized using enhanced chemiluminiscence methodand images were collected and analyzed with a Gene Genome appara-tus using Gene Snap and Gene Tools software (Syngene, Cambridge,UK).

DRG cultures

Coverslips placed in a 24wells platewere coatedwith 10 μl/ml poly-D-lysine for 2 h at 37 °C, washed, dried and further coated for 2 h with1 μl/ml laminin. DRG were harvested from adult female mice treatedwith BTZ during 1°week and from untreated mice as control. DRGwere placed in Gey's salt solution (Sigma) with 6 mg/ml glucose. Afterpeeling the DRG from connective tissue, samples were dissociated in10% collagenase, 10% trypsin and 10% DNAse in Hanks without Ca–Mgfor 30 min at 37 °C, with shaking every 10°min. Cells were resuspendedin adult Neurobasal Amedium (Gibco) supplementedwith 1× B-27, 1×penicillin/streptomycin, 1 mM L-glutamine and 6 mg/ml glucose andplated on 24 wells plates for 24°h.

For immunocytochemistry, wells were fixed in 4% paraformalde-hyde for 20°min, blocked with 1.5% normal donkey antiserum(Millipore) and incubated 2°h with antibodies anti β-tubulin-type III(1:1000, Covance), anti TNF-α (1:200, R&D Systems) or anti neurofila-ment heavy chain (1:1000, Millipore), followed by 1°h in secondaryantibody.

Fig. 1. Expression of cytokines TNF-α, IL-6, TGF-β1 and IL-1β during BTZ administration.After one week of treatment, TNF-α expression showed the highest increase, followedby a peak of IL-6 and TGF-β1 expression at twoweek. At later time points, levels decreasedbut were still higher than control ones. IL-1β expression levels increased progressivelyuntil the end of the treatment. Values expressed as mean and SEM.

167A. Alé et al. / Experimental Neurology 253 (2014) 165–173

In vivo treatment with antibodies against IL-6 and TNF-α

BTZ was given at 1 mg/kg subcutaneously twice per week for6°weeks (on days 1–4–8–11–15–18–22–25–29–32–36–39) (Brunaet al., 2010). Mice were distributed in the following groups accordingto the administered treatments:

- Group BTZ (n = 15) received only BTZ treatment and vehicle, andwas designed as a reference group

- Group anti-IL-6 (n = 13) received the samedosage of BTZ and 1 mgintraperitoneally once per week of anti-mouseIL-6 (fully rat mono-clonal antibody specific for mouse IL-6, ref MAB406 from R&DSystems, Minneapolis, MN) (Liang et al., 2006, 2009).

- Group anti-TNF-α (n = 13) received the same dosage of BTZ and0.1 mg intraperitoneally twice per week of anti-mouse TNF-α (rat/mouse chimeric antibody specific for mouse TNF, ref CNTO5048from Janssen R&D, Radnor, PA) (Charles et al., 2009; Vincent et al.,2010).

- Group control (n = 5) was constituted by untreated mice.

Functional evaluation

The functional state of the peripheral nerve was assessed by nerveconduction, rotarod and algesimetry tests. Functional evaluationswere performed at baseline, before starting treatments, and then at 4and 6°weeks of treatment.

For nerve conduction studies, the sciatic nerve was stimulatedthrough a pair of needle electrodes percutaneously placed at the sciaticnotch (proximal site) and at the ankle (distal site). Rectangular electri-cal pulses (Grass S88 stimulator) of 0.01 msdurationwere applied up to25% above the voltage that gave a maximal response. The compoundmuscle action potential (CMAP) was recorded from the plantarinterosseus muscle with microneedle electrodes. Similarly, the sensorycompound nerve action potential (SNAP) was recorded by electrodesplaced at the fourth toe near the digital nerves (Bruna et al., 2010;Navarro et al., 1994; Verdu et al., 1999). The onset latency and themaximal amplitude of the action potentials were measured and thenerve conduction velocity (NCV) of motor and sensory nerve fiberswas calculated. During electrophysiological tests, the animals wereunder anesthesia (pentobarbital 40 mg/kg i.p.) and placed on awarmedflat surface to maintain body temperature.

A rotarod apparatus for small rodents (LIAP) was used to assessgeneral sensory-motor function. Mice were placed on the rod, turningat 8 rpm, and the time that each animal remained on it before fallingwasmeasured. The value for a test was themean of three trials separat-ed by 10°min resting intervals. The ability to remain on the rotarod for120°s was taken as an index of normality (Verdu et al., 1999).

The plantar algesimetry technique was used to evaluate nociceptiveC fibers function. Mice were placed into a plastic box with an elevatedglass floor (Plantar Algesimeter, Ugo Basile) and the light of a projectionlamp was focused onto the plantar surface of one hindpaw. The time towithdraw the heated paw was obtained from a time-meter coupledwith infrared detectors. The value for a test was themean of three trialsseparated by 10 min of resting periods.

Histological methods

At the end of the treatment animals were deeply anaesthetized andintracardially perfused with paraformaldehyde (4% in PBS). A segment ofthe sciatic nerve at mid-thigh was removed. The samples were fixed inglutaraldehyde–paraformaldehyde (3%:3%), washed in cacodylatebuffer, post-fixed with 2% osmium tetroxide, dehydrated in gradedconcentrations of ethanol and embedded in epon. Light microscopyobservations were performed on 0.5 μm semithin sections stainedwith toluidine blue. Myelinated fiber countsweremade from systemat-ically selected fields at 1000× final magnification using Image J software

(NIH, Bethesda, MA). The density of myelinated nerve fibers was thenderived (Bruna et al., 2010; Gómez et al., 1996).

Plantar pads removed from perfused mice after the end of treat-ments were postfixed in Zamboni's fixative overnight and thereaftercryoprotected. Cryotome sections 70 μm thick were washed free-floating in PBS with 0.3% Triton-X100 and 1% normal goat serum for1 h, then incubated in primary rabbit antiserum against protein geneproduct 9.5 (PGP9.5, 1:1000; Ultraclone). After washes, sections wereincubated in secondary antiserum conjugated to cyanine 3 (Navarroet al., 1995). The sections were mounted on gelatin-coated slides andviewed in an epifluorescencemicroscope (Olympus BX51) using appro-priate filters. Five sections from each sample were used to quantify thenumber and density of nerve fibers present in the epidermis of the pawpads.

DRG removed from other mice after 1 and 2°weeks of treatmentwere fixed in 4% paraformaldehyde and cryoprotected. Cryotomesections 10 μm thick were washed free-floating in PBS with 0.3%Triton-X100 and 1% normal goat serum for 1 h, then incubated inprimary rabbit antiserum against β-tubulin-type III (1:1000, Covance)or anti TNF-α (1:200, R&D Systems). After washes, sections wereincubated in secondary antiserum conjugated to cyanine 3, and viewedunder epifluorescence microscopy.

Data analysis

Results are expressed as mean and standard error. The results offunctional tests during the study period are expressed as the percentagewith respect to baseline values for each mouse. Comparisons betweenexperimental groups in functional tests were made using repeatedmeasures ANOVA, whereas one-way ANOVA was used to assess differ-ences between groups in histological and immunohistochemical results.Statistical comparisons for real-time PCR resultsweremade by one-wayANOVA. Bonferroni test was used as the post hoc test when needed. A pvalue less than 0.05 was considered as significant.

Results

Proinflammatory cytokines expression during BTZ treatment

We found a high increase in mRNA expression of several cytokinesincluding TNF-α, IL-6, TGF-β and IL-1B in DRG during the 6°weeks ofBTZ treatment (Fig. 1). Just one week after treatment DRG showed thehighest expression levels of TNF-α (15.5 ± 1.5 fold increase). Thisincrease was followed by a peak of IL-6 expression (22.0 ± 6.6) andTGF-β (7.2 ± 1.6) at the second week. At later time points, levelsdecreased but were still higher than control ones, thus suggestive ofa sustained overexpression (fold increase of 4.4 ± 1.3 for TNF-α,

168 A. Alé et al. / Experimental Neurology 253 (2014) 165–173

4.6 ± 0.8 for IL-6 and 6.8 ± 1.8 for TGF-β1 at the end of the treatment).IL-1β expression levels increased progressively until the end of thetreatment (9.9 ± 3.5). Because we found the fastest and highest in-crease of TNF-α and IL-6,we also studiedmRNA levels of both cytokinesin other tissues. Interestingly, in sciatic nerve and spinal cord samplesmRNA levels of both cytokines were not or only slightly increased(Figs. 2A,B). mRNA levels of the cytokine receptors TNFR-1 and IL-6Rαfollowed the same pattern of increase observed for the cytokines inDRG, but peak expression was one week later than the peak of thecorresponding cytokine expression. In the other tissues analyzed nodifferences were observed in both cytokine receptors (Figs. 2C,D).

To localize which cells were overexpressing the cytokines, weperformed immunohistochemistry of DRG at different time points.High immunoreactivity against TNF-α was observed in neurons,particularly in small ones (Fig. 3A) at one week, that slightly decreasedat two weeks. Satellite cells, labeled against GFAP, did not show anincreased expression of TNF-α compared to controls (data not shown).In dissociated DRG cultures from animals treated with BTZ for 4 or6°weeks, TNF-α was mainly localized in the neuronal soma, as well asin the swellings of neurites (Fig. 3B).

Effects of BTZ on NF-κB

Although it was previously described that BTZ decreases TNF-α andIL-6 levels in tumoral cells, by blocking the translocation of NF-κB to thenucleus (Hideshima et al., 2001), other studies demonstrated NF-κBactivation (Hideshima et al., 2009; Li et al., 2010). Therefore, weinvestigated if NF-κB was also activated in our model, and thus explainwhy the drug increased the levels of cytokines in DRG. As we found anincrease of TNF-α already at 1°week, we analyzed at the same timepoint if p65 NF-κB subunit was translocated to the nucleus andtherefore NF-κB was activated. The p65 subunit, normally present inthe cytoplasmic fraction, was only detected in the nuclear fraction ofDRG neurons from BTZ treated animals (Fig. 4). We also observed thatp65 was phosphorylated in the cytoplasm of BTZ treated animals.Furthermore, we analyzed IκB-α mRNA expression levels and we

Fig. 2. Expression of TNF-α, IL-6 and their receptors in different nerve tissues during BTZ treatmwhereas they were slightly increased in sciatic nerve only at 2 weeks and less in spinal cord (Aexpressed as mean and SEM. *p b 0.05, **p b 0.01, ***p b 0.001 vs basal values.

observed an increase from the second to the last week of treatment,consistent with NF-κB activation (data not shown). Since the inhibitorof NF-κB-α (IκB-α) is normally degraded by the proteasome, weanalyzed the levels of this protein in the same samples, and found thatit increased, indicating that it was not degraded.

Effects of anti-TNF-α and anti-IL-6 treatment on the BTZ induced neuropathy

Groups of mice that received BTZ were treated with monoclonalantibodies against mouse TNF-α and mouse IL-6 in order to evaluate ifblockade of these cytokines might protect against the neuropathiceffects of BTZ.

Nerve conduction test resultsThe amplitude of the SNAPs in animals treated with BTZ decreased

from 2°weeks, with a reduction of 40% at 4°weeks and 50% at 6°weekswith respect to basal values. Treatment with anti-IL-6 did not signifi-cantly prevent the SNAP decrease, that was around 40% at 6°weeks.In contrast, treatment with anti-TNF-α was effective to prevent thereduction of the SNAP amplitude, which declined only 25% at the endof the treatment. Values of group anti-TNF-α were significantly higherat 4 and 6°weeks compared to group BTZ (Fig. 5A). In contrast, theCMAP amplitude of the plantar muscles was similar in all the groupsduring the follow-up (Fig. 5B).

Sensory-motor functionAt basal conditions, all animals were able tomaintain themselves on

the rotarod around 120 s (112 ± 14 s), whereas at 6°weeks, BTZ andBTZ plus anti-IL-6 groups showed a significant decline in the time ofsustained walking in the rod, with values of 87 ± 25 s for group BTZand 85 ± 35 s for group anti-IL-6. In contrast, animals treated withBTZ plus anti-TNF-α were able to maintain the position on the rodsimilar to the control group (104 ± 20 s) (Fig. 5C).

ent. Levels of both cytokines were markedly increased until the end of treatment in DRG,and B). Peak expression of TNF-α in DRG preceded IL-6 and TNFR1 peaks (C and D). Values

Fig. 3. Sections of DRG immunolabeled against β-III-tubulin (green), DAPI (blue) and TNF-α (red), from a control (top pannels) and amouse treatedwith BTZ for 1 week (bottom). TNF-αcolabels with β-III-tubulin, indicating that BTZ increases TNF-α expression in neurons, especially small ones (white arrows) (A). Dissociated primary neurons from DRG of BTZ treatedanimals for 1 week, stained against β-III-tubulin (green), DAPI (blue) and TNF-α (red). TNF-α is mainly localized in the neuronal cell body and also in neurites that presented swellings,indicating abnormal distribution of this cytokine (B). Images taken at 20×. Scale bar 50 μm (A) 50×. Scale bar 25 μm (B).

169A. Alé et al. / Experimental Neurology 253 (2014) 165–173

Pain sensibilityThermal algesimetry tests showed a progressive increase of

withdrawal thresholds in BTZ treated mice, indicative of hypoesthesia.Basal values averaged 10.4 ± 0.3 s, and were unchanged in controlmice during the follow-up. At 6°weeks of treatment the withdrawallatency increased in group BTZ (12.69 ± 0.66 s), whereas groups anti-IL-6 and anti-TNF-α maintained normal latencies (10.83 ± 1.17 s and9.83 ± 0.97 s, respectively) (Fig. 5D).

Histopathological resultsThe sciatic nerves of all the BTZ-treated mice showed a well

preserved architecture. Some signs of Wallerian degeneration wereobserved, but without marked infiltration of macrophages. In groupBTZ there was a lower density of myelinated fibers in some regions ofthe transverse sections compared to control nerves. Animals treatedwith BTZ had a reduced number of myelinated fibers (3548 ± 55) inthe sciatic nerve compared to control values (4054 ± 157), whereasin the groups treated with anti-cytokines, the number of myelinatedfibers was higher than in the BTZ treated group, although only signifi-cantly for group anti-TNF-α (3837 ± 292) (Fig. 6 and Table 2).

Skin innervationThe number of free nerve endings in the epidermis of plantar pads,

labeled for PGP9.5, was significantly lower in BTZ and anti IL-6 groupsthan in anti-TNF-α and control groups (p = 0.01). In contrast, groupanti-TNF-α did not show differences in the number of epidermalnerve fibers with respect to the control group (Fig. 7 and Table 2).

mRNA expression of proinflammatory cytokines after treatmentsAt the end of the treatment, lumbar DRG were removed to analyze

mRNA levels of TNF-α and IL-6. Anti-TNF-α treatment did not modifythe increased expression of TNF-α induced by BTZ, but reduced thelevels of IL-6 (82.2%), IL6Rα (83.2%) and TNFR1 (67.4%), whereas anti-IL-6 was able to decrease IL-6 and its receptor levels (75.5% and 83.8%respectively) but not the expression of TNF-α and TNFR1 (Fig. 8).

Discussion

The results of this study show that the proinflammatory cytokinesIL-6, TNF-α, IL-1β and TGF-β1 are increased after BTZ treatment.These cytokines are upregulated in the somas of axotomized sensoryneurons and in the distal stumps of injured peripheral nerves (Finchet al., 1993; Myers et al., 2006) and are related with a double role,neuroprotector as a first response to the stress (Figiel, 2008) andneurodegenerative because of chronic maintenance of high levels ofcytokine expression, (Miller et al., 2009; Myers et al., 2006). In ourstudy, TNF-α showed the faster increase in DRG neurons, as can beexpected of a cytokine that orchestrates the expression of otherproinflammatory cytokines, among them IL-6, that showed the higherincrease (Miller et al., 2009; Myers et al., 2006). Therefore we evaluatedthe role of both cytokines, TNF and IL-6 in BTZ toxicity. We observedthat co-administration of antibodies anti-TNF-α, but not anti-IL-6,exerts a protective effect against BTZ induced peripheral neuropathyin mice.

Fig. 4. Western blot of DRG extracts from control animals and animals treated with BTZ for 1 week. After subcellular fractionation, we found an increase of phosphorylated p65 onthe cytoplasm of BTZ treated animals (A), and translocation of p65 to the nucleus, indicative of NF-κB activation (B). IκB-α (C) and ubiquitinated proteins (D) also increased after BTZtreatment, compatible with proteasome impairment.

170 A. Alé et al. / Experimental Neurology 253 (2014) 165–173

Indeed, molecular profiling (single-nucleotide polymorphismgenotyping) has linked proinflammatory genes, included TNF-α, withBTZ induced neuropathy (Broyl et al., 2010; Favis et al., 2011). Theselectivity of a pure sensory neuropathy without motor affectation(Bruna et al., 2010, 2011), suggests that the toxic mainly affects DRG

Fig. 5. Mean amplitude of the digital nerve SNAP (A) and the plantar muscle CMAP (B). Valutreatments partially prevented the decrease of SNAP amplitude caused by BTZ, but only signifithe rotating rod relative to baseline (C). Algesimetry test results expressed as time towithdrawand reverted the hypoalgesia, compared to BTZ and BTZ + anti-IL-6 groups. *p b 0.05, **p b 0

neurons, and not the peripheral nerve. In fact, when analyzing howBTZ influenced the expression of TNF-α and IL-6, we observed anincreased expression of their levels in DRG, that wasmaintained duringall the treatment, but only a slight increase in the spinal cord and sciaticnerve. Similarly, levels of TNF-α receptor were only increased in DRG.

es are expressed as percentages with respect to baseline. Both anti-TNF-α and anti-IL-6cantly in the anti-TNF-α group. Rotarod test results, expressed as time of maintenance inal from hot pain stimulation. (D). Anti-TNF-α treatment preserved sensory-motor function.01 vs BTZ group.

Fig. 6. Semithin sections of sciatic nerves from a control mouse (A), after 6 weeks of treatment with BTZ (D), co-treatment with anti-TNF-α (B) or anti-IL-6 (C). The decreased density ofmyelinated fibers caused by BTZ was prevented with anti-TNF-α treatment. Images taken at 100×. Scale bar 20 μm.

171A. Alé et al. / Experimental Neurology 253 (2014) 165–173

All together, these data strongly suggest that sensory neurons are moreaffected than Schwann cells by BTZ. In fact, it is reported that mitochon-dria of Schwann cells have lower sensitive to chemotherapy toxicitythan neuronal mitochondria (Xiao et al., 2011; Zheng et al., 2012),thus reinforcing the hypothesis that BTZ or the consequences of protea-some inhibition mainly affect primary sensory neurons.

We observed an early peak of TNF-α overexpression after only twodoses of BTZ (one week of treatment), whereas overexpression of IL-6appeared later with a peak at two weeks (Figs. 1 and 2). This increaseof cytokines in DRG is, somehow, surprising. Inhibition of NF-κB activityis often cited as amajor contributor to the antitumoral activity of BTZ, byaffecting p65 NF-κB subunit translocation to the nucleus throughproteasome inhibition (Goldberg, 2012). This blockage can explain thereduced levels of IL-6 and TNF-α observed in tumoral cells and thedecrease of TNFRs in multiple myeloma cells after BTZ treatment(Chauhan et al., 1996; Wang et al., 1998). However, latest studiesreported controversy about NF-κB inhibition after BTZ treatment(Hideshima et al., 2009; Li et al., 2010). Interestingly, we found thatNF-κB was activated in DRG after one week of treatment, a fact that

Table 2Estimated counts of myelinated nerve fibers of the sciatic nerve at the midthigh and ofunmyelinated intraepidermal fibers in the paw skin. Partial preservation of myelinatedfibers and maintenance of intraepidermic profiles can be observed in the animalsreceiving BTZ co-treated with antibody anti TNF-α in comparison with those receivingonly bortezomib.

Group Myelinated fibers Intraepidermic fibers

CNTRL 4054 ± 157 21 ± 2.0BTZ 3548 ± 55* 11 ± 5.8**Anti-TNF-α 3837 ± 292# 20 ± 3.7##

Anti-IL-6 3792 ± 475 13 ± 5.1

Values are expressed as mean and SD. #p b 0.05, ##p b 0.01 vs BTZ group, *p b 0.05,**p b 0.01 vs control values.

may lead to the significant increase of cytokines in sensory neurons.On the other hand, we observed an increase of the IκB-α levels, suggest-ing that NF-κB is activated by the atypical way, independently of thepresence of its natural inhibitor, that it is not degraded (Mincheva-Tasheva and Soler, 2013) probably due to the impairment of theproteasome.

Since both satellite cells and sensory neurons are present in DRG, weperformed immunohistochemistry of slices and of dissociated primarycultures from DRG of treated animals to further study which cellswere expressing TNF-α. We found that the increased expression ofTNF-α induced by BTZ was mainly localized in neuronal somas, andnot in satellite cells. In agreement with our findings, expression andsecretion of TNF-α from neurons have been described after peripheralnerve and spinal cord injuries, stroke and neurodegenerative diseases(Dubovy et al., 2006; Myers et al., 2006). Therefore, the early up-regulation of TNF-α in primary sensory neurons of BTZ treated animalscould trigger the neurotoxicity, and although decreasing along thetreatment period, its increased levels may contribute to sensory neuronimpairment. The inhibition of proteasome activitymay directly alter thegene expression of neurons, mimicking signals triggered by axonaldamage of other causes (Dubovy, 2011). Since the neurons seem theprimary source of these proinflammatory cytokines, this could explainwhy immunodeficient mice still develop the neuropathy induced byBTZ (Carozzi et al., 2013).

To further corroborate the role of TNF-α in the neuropathy inducedby BTZ we studied the effects of antibodies anti-TNF-α and anti-IL-6 onthe evolution of the neuropathy. We found that anti-TNF-α treatmentbut not anti-IL-6 treatment was neuroprotective against BTZ neurotox-icity, thus indicating that TNF-α is playing a major role in this neuropa-thy. Anti-TNF-α treatment partially revert the decline of SNAPamplitude and the loss ofmyelinated fibers in the sciatic nerve, whereascompletely preserved unmyelinated epidermal nerve fibers in the skinof BTZ treated mice. These observations suggest that small nerve fibers

Fig. 7. Confocal images of skin pads immunolabeled against PGP 9.5. Representative images from the hindpaw of a control mouse (A), anti-TNF-α (B), anti-IL-6 (C) and BTZ (D) treatedmice. The density of intraepidermal profiles was higher with anti-TNF-α treatment compared with BTZ treatment alone. Images taken at 20×. Scale bar 50 μm.

172 A. Alé et al. / Experimental Neurology 253 (2014) 165–173

could be more dependent on the TNF-α pathway than largemyelinatedones. In fact, when analyzing by immunohistochemistry the DRG of BTZtreatedmice, small diameter neurons showed higher expression of TNF-α compared to large neurons.

Anti-TNF-α treatment did not affect TNF-α upregulation induced byBTZ but reversed the increased expression of TNFR1, thus suggestingthat this receptor is mediating, at least in part, the toxic effect of TNF-α. TNF-α is an early trigger of the inflammatory response (K.-M. Leeet al., 2009). The present results indicate that blockade of TNF-α doesnot only block the TNF-α pathway but also the expression of otherproinflammatory cytokines, including IL-6 and its receptor. In contrast,anti-IL-6 treatment did not show a relevant neuroprotective effect,even if it was able to reduce the upregulation of its receptor. Probably,the failure of anti-IL-6 treatment to downregulate TNFR1 explains thislack of neuroprotective effect. Similarly, co-administration of anti-IL-6and BTZ in a randomized clinical trial did not result in less neuropathythan co-administration of BTZ plus placebo (Orlowski et al., 2012).However, the fact that co-administration of anti-IL-6 monoclonalantibody did not prevent BTZ-induced neuropathy, does not meanthat IL-6 does not play a role in the neuropathy. Indeed, IL-6 receptorcan present intracrine/autocrine stimulation by its ligand (Albertiet al., 2004; Keller and Ershler, 1995), a pathway thatmakes a therapeu-tic antibody approach difficult.

Fig. 8. Expression of TNF-α, IL-6 and their receptors in DRG at the end of the co-treatments. Anti-TNF-α treatment downregulated the levels of TNFR1, IL-6 and IL6Rαexpression, whereas anti-IL-6 treatment only reduced its own expression and its receptor.*p b 0.05, ***p b 0.001 vs BTZ group.

In conclusion, our findings indicate that BTZ administration inducesa chronic increased expression of the proinflammatory cytokines TNF-αand IL-6 in DRG neurons. TNF-α seems to play a major role in theneurotoxicity of BTZ because its blockade is neuroprotective, whereasblockade of IL-6 caused limited beneficial effects on BTZ inducedneuropathy. Therefore, co-administration of anti-TNF-α in BTZ therapymay be a promising strategy to prevent the development of neuropathiccomplications. However, treatment with antibodies against TNF-α isunder discussion in cancer patients, because it is not clear if TNF-αcould increase or inhibit tumorigenesis (Williams, 2008). Our novelfindings are consistentwith other evidences thatmonoclonal antibodiesdirected against TNF-α may be suitable therapeutic agents in neuralinjuries to achieve neuroprotection and neurorepair (Sharma andSharma, 2012; Yamakawa et al., 2011). To further elucidate theetiopathogenesis of the neuropathy, studies to investigate the role ofTNF-α pathway and how BTZ could differentially regulate NF-κB intumoral and neuronal cells are needed.

Acknowledgments

The authors thank the expertise of Dr. Víctor Yuste and Dr. ElisendaCasanelles, and the technical help of Monica Espejo and JessicaJaramillo. We also acknowledge Janssen Research and Development,andMillennium Pharmaceuticals, Inc., Cambridge, MA, USA, for fundingcontribution and provision of bortezomib, and Janssen R&D, Radnor, PAUSA for provision of anti-TNF-α and anti-IL-6 antibodies. This researchwas also supported by grant PI1100464 and TERCEL and CIBERNEDfunds from the Instituto de Salud Carlos III of Spain.

References

Alberti, L., Thomachot, M.C., Bachelot, T., Menetrier-Caux, C., Puisieux, I., Blay, J.Y., 2004. IL-6as an intracrine growth factor for renal carcinoma cell lines. Int. J. Cancer 111, 653–661.

Badros, A., Goloubeva, O., Dalal, J.S., Can, I., Thompson, J., Rapoport, A.P., Heyman, M., Akpek,G., Fenton, R.G., 2007. Neurotoxicity of bortezomib therapy in multiple myeloma: asingle-center experience and review of the literature. Cancer 110, 1042–1049.

Broyl, A., Corthals, S.L., Jongen, J.L., van der Holt, B., Kuiper, R., de Knegt, Y., van Duin, M., elJarari, L., Bertsch, U., Lokhorst, H.M., Durie, B.G., Goldschmidt, H., Sonneveld, P., 2010.Mechanisms of peripheral neuropathy associated with bortezomib and vincristine inpatients with newly diagnosed multiple myeloma: a prospective analysis of datafrom the HOVON-65/GMMG-HD4 trial. Lancet Oncol. 11, 1057–1065.

Bruna, J., Udina, E., Ale, A., Vilches, J.J., Vynckier, A., Monbaliu, J., Silverman, L., Navarro, X.,2010. Neurophysiological, histological and immunohistochemical characterization ofbortezomib-induced neuropathy in mice. Exp. Neurol. 223, 599–608.

173A. Alé et al. / Experimental Neurology 253 (2014) 165–173

Bruna, J., Ale, A., Velasco, R., Jaramillo, J., Navarro, X., Udina, E., 2011. Evaluation of pre-existing neuropathy and bortezomib retreatment as risk factors to develop severeneuropathy in a mouse model. J. Peripher. Nerv. Syst. 16, 199–212.

Carozzi, V.A., Renn, C.L., Bardini, M., Fazio, G., Chiorazzi, A., Meregalli, C., Oggioni, N., Shanks,K., Quartu, M., Serra, M.P., Sala, B., Cavaletti, G., Dorsey, S.G., 2013. Bortezomib-inducedpainful peripheral neuropathy: an electrophysiological, behavioral, morphological andmechanistic study in the mouse. PLoS One 8, e72995.

Casafont, I., Berciano, M.T., Lafarga, M., 2010. Bortezomib induces the formation of nuclearpoly(A) RNA granules enriched in Sam68 and PABPN1 in sensory ganglia neurons.Neurotox. Res. 17, 167–178.

Charles, K.A., Kulbe, H., Soper, R., Escorcio-Correia, M., Lawrence, T., Schultheis, A.,Chakravarty, P., Thompson, R.G., Kollias, G., Smyth, J.F., Balkwill, F.R., Hagemann, T.,2009. The tumor-promoting actions of TNF-alpha involve TNFR1 and IL-17 in ovariancancer in mice and humans. J. Clin. Invest. 119, 3011–3023.

Chauhan, D., Uchiyama, H., Akbarali, Y., Urashima, M., Yamamoto, K., Libermann, T.A.,Anderson, K.C., 1996. Multiplemyeloma cell adhesion-induced interleukin-6 expressionin bone marrow stromal cells involves activation of NF-kappa B. Blood 87, 1104–1112.

Dubovy, P., 2011.Wallerian degeneration and peripheral nerve conditions for both axonalregeneration and neuropathic pain induction. Ann. Anat. 193, 267–275.

Dubovy, P., Jancalek, R., Klusakova, I., Svizenska, I., Pejchalova, K., 2006. Intra- andextraneuronal changes of immunofluorescence staining for TNF-alpha and TNFR1 inthe dorsal root ganglia of rat peripheral neuropathic pain models. Cell. Mol. Neurobiol.26, 1205–1217.

Favis, R., Sun, Y., van de Velde, H., Broderick, E., Levey, L., Meyers, M., Mulligan, G.,Harousseau, J.-L., Richardson, P.G., Ricci, D.S., 2011. Genetic variation associatedwith bortezomib-induced peripheral neuropathy. Pharmacogenet. Genomics 21,121–129.

Figiel, I., 2008. Pro-inflammatory cytokine TNF-alpha as a neuroprotective agent in thebrain. Acta Neurobiol. Exp. (Wars) 68, 526–534.

Finch, C.E., Laping, N.J., Morgan, T.E., Nichols, N.R., Pasinetti, G.M., 1993. TGF-beta 1 is anorganizer of responses to neurodegeneration. J. Cell. Biochem. 53, 314–322.

Fisher, R.I., Bernstein, S.H., Kahl, B.S., Djulbegovic, B., Robertson, M.J., de Vos, S., Epner, E.,Krishnan, A., Leonard, J.P., Lonial, S., Stadtmauer, E.A., O'Connor, O.A., Shi, H., Boral,A.L., Goy, A., 2006. Multicenter phase II study of bortezomib in patients with relapsedor refractory mantle cell lymphoma. J. Clin. Oncol. 24, 4867–4874.

Goldberg, A.L., 2012. Development of proteasome inhibitors as research tools and cancerdrugs. J. Cell Biol. 199, 583–588.

Gómez, N., Cuadras, J., Batí, M., Navarro, X., 1996. Histologic assessment of sciatic nerveregeneration following resection and graft or tube repair in the mouse. Restor.Neurol. Neurosci. 10, 187–196.

Hideshima, T., Chauhan, D., Schlossman, R., Richardson, P., Anderson, K.C., 2001. The roleof tumor necrosis factor alpha in the pathophysiology of human multiple myeloma:therapeutic applications. Oncogene 20, 4519–4527.

Hideshima, T., Ikeda, H., Chauhan, D., Okawa, Y., Raje, N., Podar, K., Mitsiades, C., Munshi,N.C., Richardson, P.G., Carrasco, R.D., Anderson, K.C., 2009. Bortezomib induces canonicalnuclear factor-kappaB activation in multiple myeloma cells. Blood 114, 1046–1052.

Keller, E.T., Ershler, W.B., 1995. Effect of IL-6 receptor antisense oligodeoxynucleotide onin vitro proliferation of myeloma cells. J. Immunol. 154, 4091–4098.

Lee, H.K., Shin, Y.K., Jung, J., Seo, S.-Y., Baek, S.-Y., Park, H.T., 2009. Proteasome inhibitionsuppresses Schwann cell dedifferentiation in vitro and in vivo. Glia 57, 1825–1834.

Lee, K.-M., Jeon, S.-M., Cho, H.-J., 2009. Tumor necrosis factor receptor 1 induces interleukin-6 upregulation through NF-kappaB in a rat neuropathic pain model. Eur. J. Pain 13,794–806.

Leung, L., Cahill, C.M., 2010. TNF-alpha and neuropathic pain—a review.J. Neuroinflammation 7, 27.

Li, C., Chen, S., Yue, P., Deng, X., Lonial, S., Khuri, F.R., Sun, S.-Y., 2010. Proteasome inhibitorPS-341 (bortezomib) induces calpain-dependent IkappaB(alpha) degradation. J. Biol.Chem. 285, 16096–16104.

Liang, B., Gardner, D.B., Griswold, D.E., Bugelski, P.J., Song, X.Y., 2006. Anti-interleukin-6monoclonal antibody inhibits autoimmune responses in a murine model of systemiclupus erythematosus. Immunology 119, 296–305.

Liang, B., Song, Z., Wu, B., Gardner, D., Shealy, D., Song, X.Y.,Wooley, P.H., 2009. Evaluationof anti-IL-6 monoclonal antibody therapy using murine type II collagen-inducedarthritis. J. Inflamm. (Lond.) 6, 10.

Mateos, M.-V., Hernandez, J.M., Hernandez, M.T., Gutierrez, N.C., Palomera, L., Fuertes, M.,Garcia-Sanchez, P., Lahuerta, J.J., de la Rubia, J., Terol, M.-J., Sureda, A., Bargay, J., Ribas,P., Alegre, A., de Arriba, F., Oriol, A., Carrera, D., Garcia-Larana, J., Garcia-Sanz, R., Blade,J., Prosper, F., Mateo, G., Esseltine, D.-L., van de Velde, H., San Miguel, J.F., 2008.Bortezomib plus melphalan and prednisone in elderly untreated patients withmultiple myeloma: updated time-to-events results and prognostic factors for timeto progression. Haematologica 93, 560–565.

Miller, R.J., Jung, H., Bhangoo, S.K., White, F.A., 2009. Cytokine and chemokine regulationof sensory neuron function. Handb. Exp. Pharmacol. 417–449.

Mincheva-Tasheva, S., Soler, R.M., 2013. NF-kappaB signaling pathways: role in nervoussystem physiology and pathology. Neuroscientist 19, 175–194.

Myers, R.R., Shubayev, V.I., 2011a. The ology of neuropathy: an integrative review of therole of neuroinflammation and TNF-alpha axonal transport in neuropathic pain.J. Peripher. Nerv. Syst. 16, 277–286.

Myers, R.R., Campana, W.M., Shubayev, V.I., 2006. The role of neuroinflammationin neuropathic pain: mechanisms and therapeutic targets. Drug Discov. Today 11, 8–20.

Navarro, X., Verdu, E., Buti, M., 1994. Comparison of regenerative and reinnervatingcapabilities of different functional types of nerve fibers. Exp. Neurol. 129, 217–224.

Navarro, X., Verdu, E., Wendelscafer-Crabb, G., Kennedy, W.R., 1995. Innervation of cutane-ous structures in the mouse hind paw: a confocal microscopy immunohistochemicalstudy. J. Neurosci. Res. 41, 111–120.

Orlowski, R.Z., Gercheva, L., Williams, C., Sutherland, H.J., Robak, T., Masszi, T., Goranova-Marinova, V., Dimopoulos, M.A., Cavenagh, J.D., Spicka, I., Maiolino, A., Suvorov, A.,Blade, J., Samoilova, O.S., Van De Velde, H., Bandekar, R., Kranenburg, B., Xie, H.,Rossi, J.-F., 2012. Phase II, randomized, double blind, placebo controlled studycomparing siltuximab plus bortezomib versus bortezomib alone in patients withrelapsed/refractory multiple myeloma. J. Clin. Oncol. 30, 8018 (ASCO MeetingAbstracts).

Palanca, A., Casafont, I., Berciano, M.T., Lafarga, M., 2013 (in press). Proteasome inhibitioninduces DNA damage and reorganizes nuclear architecture and protein synthesismachinery in sensory ganglion neurons. Cell. Mol. Life Sci.

Poruchynsky, M.S., Sackett, D.L., Robey, R.W., Ward, Y., Annunziata, C., Fojo, T., 2008.Proteasome inhibitors increase tubulin polymerization and stabilization in tissueculture cells: a possible mechanism contributing to peripheral neuropathy andcellular toxicity following proteasome inhibition. Cell Cycle 7, 940–949.

Richardson, P.G., Barlogie, B., Berenson, J., Singhal, S., Jagannath, S., Irwin, D., Rajkumar,S.V., Srkalovic, G., Alsina, M., Alexanian, R., Siegel, D., Orlowski, R.Z., Kuter, D.,Limentani, S.A., Lee, S., Hideshima, T., Esseltine, D.-L., Kauffman, M., Adams, J.,Schenkein, D.P., Anderson, K.C., 2003. A phase 2 study of bortezomib in relapsed,refractory myeloma. N. Engl. J. Med. 348, 2609–2617.

Rockwell, P., Yuan, H.M., Magnusson, R., Figueiredo-Pereira, M.E., 2000. Proteasomeinhibition in neuronal cells induces a proinflammatory response manifested byupregulation of cyclooxygenase-2, its accumulation as ubiquitin conjugates, andproduction of the prostaglandin PGE(2). Arch. Biochem. Biophys. 374, 325–333.

SanMiguel, J.F., Schlag, R., Khuageva, N.K., Dimopoulos, M.A., Shpilberg, O., Kropff, M., Spicka,I., Petrucci, M.T., Palumbo, A., Samoilova, O.S., Dmoszynska, A., Abdulkadyrov, K.M.,Schots, R., Jiang, B., Mateos, M.-V., Anderson, K.C., Esseltine, D.L., Liu, K., Cakana, A., vande Velde, H., Richardson, P.G., 2008. Bortezomib plus melphalan and prednisone forinitial treatment of multiple myeloma. N. Engl. J. Med. 359, 906–917.

Sharma, A., Sharma, H.S., 2012. Monoclonal antibodies as novel neurotherapeutic agentsin CNS injury and repair. Int. Rev. Neurobiol. 102, 23–45.

Staff, N.P., Podratz, J.L., Grassner, L., Bader, M., Paz, J., Knight, A.M., Loprinzi, C.L., Trushina,E., Windebank, A.J., 2013. Bortezomib alters microtubule polymerization and axonaltransport in rat dorsal root ganglion neurons. Neurotoxicology 39C, 124–131.

Verdu, E., Vilches, J.J., Rodriguez, F.J., Ceballos, D., Valero, A., Navarro, X., 1999. Physiologicaland immunohistochemical characterization of cisplatin-induced neuropathy in mice.Muscle Nerve 22, 329–340.

Vincent, M., Sayre, N.L., Graham,M.J., Crooke, R.M., Shealy, D.J., Liscum, L., 2010. Evaluation ofan anti-tumor necrosis factor therapeutic in a mouse model of Niemann–Pick C liverdisease. PLoS One 5, e12941.

Wang, C.Y., Mayo, M.W., Korneluk, R.G., Goeddel, D.V., Baldwin Jr., A.S., 1998. NF-kappaBantiapoptosis: induction of TRAF1 and TRAF2 and c-IAP1 and c-IAP2 to suppresscaspase-8 activation. Science 281, 1680–1683.

Watanabe, T., Mitsuhashi, M., Sagawa, M., Ri, M., Suzuki, K., Abe, M., Ohmachi, K., Nakagawa,Y., Nakamura, S., Chosa, M., Iida, S., Kizaki, M., 2013. Phytohemagglutinin-induced IL2mRNA in whole blood can predict bortezomib-induced peripheral neuropathy formultiple myeloma patients. Blood Cancer J. 3, e150.

Williams, G.M., 2008. Antitumor necrosis factor-alpha therapy and potential cancerinhibition. Eur. J. Cancer Prev. 17, 169–177.

Xiao, W.H., Zheng, H., Zheng, F.Y., Nuydens, R., Meert, T.F., Bennett, G.J., 2011. Mitochondrialabnormality in sensory, but not motor, axons in paclitaxel-evoked painful peripheralneuropathy in the rat. Neuroscience 199, 461–469.

Yamakawa, I., Kojima, H., Terashima, T., Katagi, M., Oi, J., Urabe, H., Sanada, M., Kawai, H.,Chan, L., Yasuda, H., Maegawa, H., Kimura, H., 2011. Inactivation of TNF-alphaameliorates diabetic neuropathy in mice. Am. J. Physiol. Endocrinol. Metab. 301,E844–E852.

Zheng, H., Xiao, W.H., Bennett, G.J., 2012. Mitotoxicity and bortezomib-induced chronicpainful peripheral neuropathy. Exp. Neurol. 238, 225–234.